Supplementary Materials (Tables)

Supplementary Table 1. Fitness comparisons of unpassaged versus post-passage viruses by host strain.

Pair-wise comparison / Proviral Load (F-MuLV/GAPD copies) / Infectious Particles (FFU/spleen)
Host Genotype / Infected with… / Mean, SD, n / ANOVA / Mean, SD, n / ANOVA
A/WySn / Unpassaged / 0.005±0.004, n=7 / (F1,17=162.7, p<0.0001) / 47.14±18.8, n=7 / F1,19=965.03, p<0.0001
A/WySn passaged / 0.45±0.29,n=12 / 30562.85±15462.35, n=14
BALB/c / Unpassaged* / 0.008±0.01, n=33 / (F1,93=405.37, p<0.0001) / 139.06±258.55, n=36 / (F1,96=676.87, p<0.0001)
BALB/c passaged* / 0.58±0.47, n=61 / 15996.33±11361.22, n=61
DBA/2J / Unpassaged / 0.1±0.06, n=8 / (F1,22=71.91, p<0.0001) / 3258.33±3340.61, n=8 / (F1,22=34.96, p<0.0001)
DBA/2J passaged / 0.62±0.32, n=16 / 14796.28±5338.25, n=16

* The “unpassaged” and “BALB/c passaged” data sets are based on pooled estimates from previously published data(1).

Supplementary Table 2. Virulence comparisons of unpassaged versus post-passage viruses by host strain.

Pair-wise comparison / Splenomegaly (grams spleen)
Host Genotype / Infected with… / Mean, SD, n / ANOVA
A/WySn / Unpassaged / 0.16±0.02, n=7 / (F1,20=481.17, p<0.0001
A/WySn passaged / 2.09±0.23, n=15
BALB/c / Unpassaged* / 0.18±0.08, n=36 / (F1,96=555.28, p<0.0001)
BALB/c passaged* / 3.06±0.68, n=61
DBA/2J / Unpassaged / 1.06±0.37, n=8 / (F1,22=59.09, p<0.0001)
DBA/2J passaged / 2.08±0.27, n=16

The BALB/c data set is based on pooled data across three BALB/c MHC-congenic data sets.

*The “unpassaged” and “BALB/c passaged” data sets are based on pooled estimates from previously published data(1).

Supplementary Table 3. Effect of host familiarity on post-passage viral fitness and virulence.

Virus Genotype / Host Infected / Proviral Load / Infectious Virus Particles / Splenomegaly
A/WySn Passaged / Familiar / (F1,46=16.89, p=0.0002) / (F1,51=24.82, p<0.0001) / (F1,52=17.80, p<0.0001)
Unfamiliar
BALB/c Passaged / Familiar* / (F1,125=20.58, p<0.0001) / (F1,106=23.08, p<0.0001) / (F1,125=9.06, p=0.0032)
Unfamiliar
DBA/2J Passaged / Familiar / (F1,51=18.97, p<0.0001) / (F1,54=5.49, p=0.02) / (F1,54=2.88, p=0.09)
Unfamiliar

Summary statistics are based on the results of a one-way ANOVA comparing the means of residual variation between “familiar” and “unfamiliar” hosts infected with the same post-passage virus stock. A one-step GLM model incorporating the effects of “host genotype”, “virus genotype”, and “host familiarity” was not used due to an insufficient number of degrees of freedom per independent experiment. More specifically, the effects of “host genotype” and “virus genotype” use up all available degrees of freedom in the model containing all three effects, thus it is not possible to obtain a significance estimate for the effect of “host familiarity” after controlling for the other two variables in the same model. To overcome this limitation we employed a two-step analysis approach whereby a GLM was constructed that incorporated the effect of “host genotype” and “virus genotype” (step 1). Residual variation was then extracted from this model and was used to test for a significant effect of host familiarity on patterns of viral fitness and virulence (step 2). Units are not provided as results are based on residuals which are unit-less.

*= The “Familiar” BALB/c data set (shown in Figure 3 and Figure 4A) that was used in this analysis is based on a pooled estimate from previously published data(1), while the “Unfamiliar” BALB/c data set is unique to this study.

Supplementary Table 4. Pair-wise comparisons of viruses infecting their familiar host versus unfamiliar hosts

possessing novel MHC polymorphisms alone or novel MHC and non-MHC polymorphisms.

Pair-wise Comparison / Proviral Load / Infectious Virus Particles / Splenomegaly
Familiar
vs.
Unfamiliar (MHC) / (F1,195=31.07, p<0.0001) / (F1,200=50.42, p<0.0001) / (F1,206=16.99, p<0.0001)
Familiar
vs.
Unfamiliar (MHC and non-MHC) / (F1,227=49.90, p<0.0001) / (F1,216=50.85, p<0.0001) / (F1,236=26.90, p<0.0001)
Unfamiliar MHC
vs.
Unfamiliar MHC and non-MHC / (F1,243=6.30, p=0.01) / (F1,233=2.61, p=0.11) / (F1,257=0.72, p=0.40)

Units are not provided as statistical results are based on comparisons between mean residual variation extracted from a generalized linear model (see Material and Methods), which are unit-less values. A GLM was constructed that incorporated the effect of “host genotype” and “virus genotype”. Residual variation was then extracted from this model and was used to test for significant differences in viral fitness and virulence under the following three treatments: exposure to familiar host genotypes, exposure to hosts possessing novel MHC polymorphisms alone, and exposure to hosts possessing novel MHC and non-MHC polymorphisms.

Supplementary Table 5. Jackson Laboratories mouse strain information.

Abbreviated ID / H2 Haplotype / Full Strain ID / Catalog Number
BALB/c-H2d / d / BALB/cJ / cat#000651
BALB/c-H2k / k / C.C3-H2k / cat#001951
BALB/c-H2b / b / C.B10-H2b/LilMcdJ / cat#001952
DBA/2J-H2d / d / DBA/2J / cat#000671
A/WySn-H2s / s / A.SW-H2s H2-T18b/SnJ / cat#000471

-a more detailed description of the derivation of each of these mouse lines can be found on the Jackson Laboratories website (jaxmice.jax.org).

Supplementary Table 6. Results from a GLM analyzing the effect of host genotype on proviral load, infectious particle counts, and splenomegaly during infection with unpassaged virus.

Host Genotype by Proviral Load
DF / S.S. / Mean Square / F Ratio / p-value
Model / 4 / 82.32026 / 20.5801 / 10.9888 / <.0001
Error / 43 / 80.53166 / 1.8728
C. Total / 47 / 162.8519
Host Genotype by Infectious Particles
DF / S.S. / Mean Square / F Ratio / p-value
Model / 4 / 82.11401 / 20.5285 / 16.977 / <.0001
Error / 46 / 55.62292 / 1.2092
C. Total / 50 / 137.7369
Host genotype by Splenomegaly
DF / S.S. / Mean Square / F Ratio / p-value
Model / 4 / 5.228145 / 1.30704 / 52.0408 / <.0001
Error / 46 / 1.155318 / 0.02512
C. Total / 50 / 6.383463

Supplementary Table 7. Results from a GLM analyzing the effect of host genotype on proviral load, infectious particle counts, and splenomegaly during infection with passaged viruses.

Host Genotype by Proviral Load
DF / S.S. / Mean Square / F Ratio / p-value
Model / 4 / 89.37121 / 22.3428 / 7.637 / <.0001
Error / 223 / 652.4067 / 2.9256
C. Total / 227 / 741.7779
Host Genotype by Infectious Particles
DF / S.S. / Mean Square / F Ratio / p-value
Model / 4 / 29.89306 / 7.47326 / 2.8141 / 0.0263
Error / 212 / 562.9921 / 2.65562
C. Total / 216 / 592.8852
Host Genotype by Splenomegaly
DF / S.S. / Mean Square / F Ratio / p-value
Model / 4 / 87.49437 / 21.8736 / 26.7486 / <.0001
Error / 232 / 189.7171 / 0.8177
C. Total / 236 / 277.2115

Supplementary Table 8. Results from a GLM analyzing the effect of virus genotype on proviral load, infectious particle counts, and splenomegaly during infection with passaged viruses.

Virus Genotype by Proviral Load
DF / S.S. / Mean Square / F Ratio / p-value
Model / 8 / 53.89751 / 6.73719 / 2.1449 / 0.0329
Error / 219 / 687.8804 / 3.14101
C. Total / 227 / 741.7779
Virus Genotype by Infectious Particles
DF / S.S. / Mean Square / F Ratio / p-value
Model / 8 / 88.44365 / 11.0555 / 4.5586 / <.0001
Error / 208 / 504.4415 / 2.4252
C. Total / 216 / 592.8852
Virus Genotype by Splenomegaly
DF / S.S. / Mean Square / F Ratio / p-value
Model / 8 / 31.13065 / 3.89133 / 3.6054 / 0.0006
Error / 228 / 246.0809 / 1.0793
C. Total / 236 / 277.2115

Supplementary Table 9. Results from a GLM analyzing the effect of host-genotype x virus-genotype interactions on proviral load, infectious particle counts, and splenomegaly during Infection with post-passage virus.

H x V Interaction by Proviral Load
DF / S.S. / Mean Square / F Ratio / p-value
Model / 23 / 420.7048 / 18.2915 / 11.6219 / <.0001
Error / 204 / 321.0732 / 1.5739
C. Total / 227 / 741.7779
H x V Interaction by Infectious Particles
DF / S.S. / Mean Square / F Ratio / p-value
Model / 21 / 263.5005 / 12.5476 / 7.4284 / <.0001
Error / 195 / 329.3847 / 1.6892
C. Total / 216 / 592.8852
H x V Interaction by Splenomegaly
DF / S.S. / Mean Square / F Ratio / p-value
Model / 23 / 163.0165 / 7.08767 / 13.2201 / <.0001
Error / 213 / 114.195 / 0.53613
C. Total / 236 / 277.2115

1. Kubinak JL, Ruff JS, Hyzer CW, Slev PR, Potts WK. Experimental viral evolution to specific host MHC genotypes reveals fitness and virulence trade-offs in alternative MHC types. Proc Natl Acad Sci U S A. 2012 Feb 7.